ARTIFICIAL NEURAL NETWORK-BINARY LABEL CLASSIFICATION

Saurav Singh, MahimaParashar

Abstract: Artificial Neural Networks (ANNs) are a family of models which try to mimic the network of neurons in the brain to process information. It is a system of interconnected neurons which exchange the messages between each other. When too many classifications are involved, the ANN tends to use up a lot of memory space for bigger weight matrices and more output nodes corresponding to more classification labels. The goal of our research is to analyze the existing ANN model and develop an algorithm that can give output in binary format instead of dedicating one output node to each label. This will help in reducing the number of output nodes and weight matrices and reduce the memory space required while maintaining the accuracy of the network.

Keywords: Binary, Classification, Network, Neural, Supervised

DOI: https://doi.org/10.15623/ijret.2016.0507024